in west greenlandic...department of language, literature and media institute of culture, language...

Semantic roles of the

terminative casein West Greenlandic

– an initial study forGreenlandic-Danishmachine translation

Ataqqinartu-aqqa-l-li eqqarsaati-n-nit alla-mut saa-tip-pa-a-ngaPrince.little.REL.s.but thought.1sPOSS.ABL.p other.TRM.s turn.CAUS.IND.3s.1sO@SUBJ @ADVL @ADVL @PRED§AG §ORI §DES-LOC §PRED

((VII.26) But the Little Prince made me turn away from my thoughts to something else.)

Liv MolichValency and semantic roles, bachelor, 10 ECTS

Department of Language, Literature and MediaInstitute of Culture, Language and History

University of GreenlandJune 29th 2018

1

“ ”

Liv Molich

Valency and semantic roles, bachelor, 10 ECTSDepartment of Language, Literature and Media

Institute of Culture, Language and HistoryUniversity of Greenland

Table of ContentsIntroduction..........................................................................................................................................3Theory...................................................................................................................................................3

Semantic roles..................................................................................................................................4Semantic maps.................................................................................................................................6Terminative case..............................................................................................................................7

Mapping the terminative case...............................................................................................................9Core clause-level roles...................................................................................................................10

§REC.........................................................................................................................................10§COG........................................................................................................................................10§EXP.........................................................................................................................................11§BEN.........................................................................................................................................12§TH...........................................................................................................................................12§REFL.......................................................................................................................................13

Spatial and temporal roles..............................................................................................................14§DES.........................................................................................................................................14§LOC, §PATH and §EXT..........................................................................................................16

Other roles......................................................................................................................................17§FIN..........................................................................................................................................17§CAU........................................................................................................................................17§SOA and §TP..........................................................................................................................17§ATR, §CONT and §COMP.....................................................................................................18

Lexicalized usage...........................................................................................................................18Discussion...........................................................................................................................................19Conclusion..........................................................................................................................................19Literature............................................................................................................................................21

Theory............................................................................................................................................21Corpora..........................................................................................................................................22Tools...............................................................................................................................................22

Appendix I: Semantic maps on terminative in Greenlandic...............................................................23Appendix II: Distribution of semantic roles for the Greenlandic terminative....................................24Appendix III: Semantic roles and other abbreviations.......................................................................26

Syntactic functions.........................................................................................................................26Semantic roles................................................................................................................................26Other abbreviations........................................................................................................................27

Appendix IV: Corpus..........................................................................................................................30

2

Liv Molich



IntroductionThis paper is based on the course “Valency and semantic roles” in the spring term 2018 at University of Greenland. Its purpose, however, is not purely theoretical, as the investigation is aimed at contributing to a machine translation project initiated by The Language Secretariat of Greenland. Among the activities of this project are the development of a syntactic analyzer for Greenlandic (https://oqaasileriffik.gl/langtech/live/) and a lexical database (https://oqaasileriffik.gl/search/dictionary/). The semantic roles will be marked in the database, but so far for the Danish verbs only. In order to take advantage of these roles, it must be sorted out which roles are relevant for Greenlandic, and the information must be added to the database as well as to the syntactic analyzer. Therefore, this paper is intended to serve as the first step of a thorough investigation of the Greenlandic case system which, I hope, will improve on the quality of the machine translations between Greenlandic and Danish. My colleague, Dr. Eckhard Bick, has been kind to answer questions on semantic role marking when doubt has occurred.

My paper will focus on the terminative case in West Greenlandic. The terminative case is particularly interesting because it covers both core syntactic and spatial functions, in addition to other functions. In my investigation, I will describe the functions of the terminative case in Greenlandic and attempt to map the semantic roles of the Greenlandic terminative case on a semantic map. The study will show how the semantic map method and the semantic role theory can be used also for Greenlandic.

I will shortly explain what semantic roles are, what semantic maps are used for, and how the terminative case and its roles are described in the linguistic literature on Greenlandic. On the basis of that, I will perform an investigation of terminatives in a minor text corpus.

TheorySemantics and syntax are closely related phenomena. Therefore, the mapping of semantics and syntax will be helpful in describing languages for the use of rule based machine translation:

“Depending, for instance, on the number of obligatory arguments, several valency or semantic frames may share the same verb sense, but two different verb senses will almost always differ in at least one syntactic or semantic aspect of their argument frame – guaranteeing that all senses can in principle be disambiguated exploiting a parser's argumenttags and dependency links” (Bick 2011: 2).

Eckhard Bick uses this method for machine translation between a number of Indo-European languages, Danish included. “Even in a case-poor language like Danish, we found some clear likelihood relations between thematic roles and syntactic functions,” Bick explains (Bick 2011: 3).

3

Liv Molich



According to the modern valency theory established by Lucien Tesnière, valency is the ability of verbs to call for and bind complements in a sentence. Tesnière describes valency in a rather quantitative way, counting the complements: The verb compares to an “atom with a particular number of hooks that can – according to the number of hooks – attract a varying number of actants, which it keeps in its dependence. The number of hooks that a verb possesses, and consequently the number of actants that it governs, constitutes what we call the valency of a verb” (Tesnière 1976: 238, cf. Ágel and Fischer 2015: 232). The newer development of valency theories have brought a great deal of semantics into play, investigating the complements from a more qualitative view, looking at valency on different levels and distinguishing between them (Ágel and Fischer 2015; Hansen and Heltoft 2011: 106f.).

Unfortunately, even with its semantic layer, valency theory is not particularly useful for the current description of Greenlandic. The reason for this is that traditional descriptions of Greenlandic grammar, the above-mentioned syntactic analyzer included, describe all occurrences of the terminative case as facultative adverbials, except from a few facultative adnominals, agents and reflexives, while abstaining from linking the different valency potential of the verb to the semantics of the adverbials. A direct relation between the semantic roles and syntactic functions for Greenlandic thus cannot be established in a useful way.

However, using a combination of valency theory and transformational grammar theory – for Greenlandic developed by Karen Langgård (a brief overview of her initial studies can be seen in Langgård 1992) – the description of syntactic functions can be useful for the mapping of semantic roles on core clause-level, which also Zúñiga shows for beneficiaries (Zúñiga 2011).

Semantic rolesCharles Fillmore has developed a theory for describing language from a more semantic point of view. This theory is based on the transformational grammar and introduces deep cases, which correspond to semantic or thematic roles, in his well-known paper The Case for Case (Fillmore 1968). In modern days, semantic roles are particularly popular among computational linguists marking corpora for machine translation purposes (Jurafsky and Martin 2015: 2).

While a useful correlation between the syntactic functions and the semantic roles probably can be established for core clause-level roles – which I define as subjects and objects, whether obligatory or not – the same useful correlation probably cannot be established for the non-core clause-level roles; but also in this respect, mapping the semantic roles might be useful for translation purposes since a semantic map can show cross-linguistic variations needed for translation.

Syntactic functions change with the syntax. For example, once an active sentence has been passivized, the former object is no longer an object, but turns into a subject, and other roles change

4

Liv Molich



as well. This is interesting for syntactical and morphological reasons, but for the purpose of cross-linguisitic comparison in semantic maps, a more stable system might be useful. Semantic roles (§) do not change in the same way as syntactic functions (@). In this way, they point at an alternative method for describing the roles of a sentence. A list of abbreviations and roles can be found in Appendix III.

Danish:(D1) Anna slog Peter med en pind

Anna beat.ACT Peter stick.INDEF with.PRP@SUBJ @PRED @OBJ @ADVL§AG §PRED §PAT §INS

(Anna beated Peter with a stick.)

(D2) Peter blev slået af Anna med en pindPeter beat.PASS Anna by.PRP stick.indef with.PRP@SUBJ @PRED @AG @ADVL§PAT §PRED §AG §INS

(Peter was beaten by Anna with a stick.)

Greenlandic:(G1) Anna-p Piitaq ajaappia-mik unatar-paa

Anna.REL.s Peter.ABS.s stick.INS.s beat.3s.3sO.IND@SUBJ @OBJ @ADVL @PRED§AG §PAT §TH §PRED

(Anna beated Peter with a stick.)

(G2) Piitaq Anna-miit ajaappia-mik unatar-neqar-poqPeter.ABS.s Anna.ABL.s stick.INS.s beat.PASS.3s.IND@SUBJ @AG @ADVL @PRED§PAT §AG §TH §PRED

(Peter was beaten by Anna with a stick.)

These examples clearly show that semantic roles and syntactic functions represent two different ways of describing connections between the arguments of a sentence. The syntactic functions are defined with reference to the syntactic context and change when the syntax changes. Conversely, the semantic roles do not change. Anna stays agent, no matter if the sentence becomes passive. Furthermore, the examples indicate that semantic roles are cross-linguistically comparable. For this reason, semantic roles are particularly interesting for cross-linguistic investigations.

5

Liv Molich



For Danish, Eckhard Bick has used 38 different semantic roles (Bick 2011), while his description ofPortuguese includes 53 different roles (Bick 2007). The need for roles vary from language to language. My investigation will use Danish semantic roles, but delete roles or add more roles from the Portuguese list where the Danish list is not comprehensive as regards the description of Greenlandic semantic roles.

Semantic mapsMartin Haspelmath has developed a method for cross-linguistic comparison of semantic function in the form of semantic maps (Haspelmath 2003: 213). The maps contain a number of main functions related to a particular case or affix.

The semantic maps make it possible to empirically observe the use of a particular gram – function word and affix – in a certain language. These language-specific observations can then be drawn on the same map to show the similarities between languages (Haspelmath 2003: 238-239). In this way, mapping the semantics of different grams on the same map makes it possible to visually show the connections between different grams cross-linguistically.

The map should be used cross-linguistically because only when two languages with different function grams meet, it is possible to distinguish the different functions. In the case of two languages with similar functions, there is no reason to distinguish between the functions.

The ideal purpose of drawing semantic maps is to deduce the functions needed for all languages in the world and how they relate to each other (Haspelmath 2003: 216). This thought is, however, utopian, but the map can be drawn for a number of languages and for a number of grams. Thereby, knowledge of assumed universals can be established, and our understanding of how languages are similar to each other may become more precise. According to Haspelmath, a dozen of genealogically diverse languages would be sufficient to establish a stable map that would only need minor alterations (Haspelmath 2003: 216-217).

Haspelmath argues that a simple map with not too many connection lines between the functions is of the highest interest when looking for similarities between languages, even though there might be grams which cannot be covered fully by the map (Haspelmath 2003: 218). Such grams and functions which are not possible to fit into the map will show the diversity of language. Whichever focus a researcher has, the maps can be connected or subdivided to fit the purpose (Haspelmath 2003: 230).

For example, the dative map consists of a number of different functions typical for the dative case, such as purpose, direction, recipient, experiencer, predicative possessor, external possessor, beneficiary and judicantis. The fact that the map is cross-linguistic and universal, does, however, not mean that all functions are covered by the dative case in all languages that use the dative case.

6

Liv Molich



On the contrary, most grams cover only some of the functions, and other functions do not exist on the map (Haspelmath 2003: 214-5). The functions most likely covered by the same language are connected to each other with lines.

A semantic map for comparing French dative and à. Haspelmath 2003: 219.

The different functions of the semantic map are chosen because of their universality. Not all of themare differing on a semantic level, despite the fact that the map is called “semantic”. Some of the functions also cover syntactic information.

As regards Greenlandic, the case system has long been known and mapped generally (Malchukov and Narrog 2009: 532-533). However, a more comprehensive and fine-grained map has still not been made.

Terminative caseWest Greenlandic is a split ergative, polysynthetic language of the Eskimo-Aleut language family (for a brief introduction, see Sadock 2003). The Greenlandic verb either marks subject, object or both. When the subject or object are explicit, the “grammatical” cases absolutive and relative are used according to the ergative system in which the case of the subject changes with the inflexion of the verb.

Another six cases are oblique, and do never agree with the verb. Here, I will focus on the oblique, terminative case – sometimes also called the allative. The terminative case in West Greenlandic covers the typical dative functions of recipient of ditransitives, benefactive and purpose. I will countthe functions of recipient and benefactive as core syntactic. As regards spatial functions, goal and location are functions of the terminative (Næss 2009). For this reason, the terminative case is also called allative. Apart from this, both Michael Fortescue, Langgård and Langgård and Sadock list a not exhaustive number of other uses, which include reflexive and lexicalized constructions. Usually,

7

Liv Molich



any noun is facultative in a Greenlandic sentence (Fortescue 1984; Langgård and Langgård 1988; Sadock 2003).

However, neither Langgård and Langgård, Fortescue nor Sadock show the distribution of the terminative case in different functions. For my machine translation project, I will have to investigatethe distribution of the terminative case and to compare it to the prepositions used for the same functions in Danish. How fine-grained this distribution chart will be, depends on the number of semantic roles taken in consideration.

According to Åshild Næss, “the dative is perhaps the most difficult to define in a consistent, cross-linguistically valid way” (Næss 2009: 572). This is also true as regards the West Greenlandic terminative case which to a large extent is used like the Indo-European dative case. The semantic map of the dative case can therefore be used for my investigation of the terminative case.

Even though the terminative case is clearly distinguishable, by the inflexional morpheme TRM which appears in four different homosemes, -mut, -nut, -unnga, or -nunnga, one can not be sure thatthe terminative case is really a terminative since studies show that case polysemy often occurs for both phonetic and semantic reasons (Malchukov and Heiko 2009).

A widely known example is the transitive verb love, asa-, which has an oblique object when it becomes monocongruent by the derivational, “half-transitivizing” morpheme. This object is undergoing a shift from the traditional instrumental case to the terminative case:

(G3) Meeqqa-nik asa-nnin-neru-vo-q Child.INS.p love.HTR.COMP.IND.3s@ADVL @PRED§TH §PRED

(He loves children more; Kalaallit Nunaata Radioa 2005)

(G4) Angum-mut alla-mut asa-nni-ler-po-qMan.TRM other.TRM love.HTR.INCH.IND.3s@ADVL @N-CONG @PRED§TH (§TH) §PRED

(She has begun to love another man; Kalaallit Nunaata Radioa 2007)

Næss explains that dative-accusative and dative-allative syncretism are very frequent, and also dative-locative and dative-genitive syncretism are regularly found (Næss 2009: 580). Similar kinds of syncretism are found in Greenlandic, too, which is indicated by a wide range of designations of the cases (e.g. Sadock 2003: 6 vs. Nielsen 2014: 33-34, 36). This could explain the wide range of functions attached to the Greenlandic terminative case.

8

Liv Molich



Mapping the terminative caseFor the purpose of investigating the functions of the terminative case, a semantic map will be useful. From that the functions can easily be seen and compared to the prepositions used for similar expressions in Danish.

The semantic map is very useful for comparing languages (Haspelmath 2003: 214). This also countsfor Greenlandic. As the terminative case in Greenlandic covers most functions of the dative map, my aim is to show more precisely which typical dative functions are covered by the terminative case, when the terminative is used, and which functions are the most frequent.

The text chosen for the investigation is the first fifteen chapters of Le Petit Prince written by Antoine de Saint-Exupéry, in the Greenlandic translation (Saint-Exupéry 2017). The Greenlandic translation of the text is very close to the Danish one, which make them easily comparable semantically, even though the morphology and syntax of the languages are very dissimilar. The text is widely used for cross-linguistic investigations, so my results can be used for cross-linguistic comparisons with other languages as well.

My method will be to extract all sentences containing terminative case marking, and manually go through the examples to map the semantic functions. The sentences can be found in Appendix IV, both in Greenlandic and in Danish. The Greenlandic examples used in my investigation are provided with English translations.

I will mark the syntactic functions as well as the semantic roles in the examples. The dependencies are not marked, but can to some extent be found through the live analyzer for Greenlandic (https://oqaasileriffik.gl/langtech/live/). However, these analyses are not completely reliable before the end of 2018, when I have implemented my results from my ongoing investigation of terminatives in the tool.

In the first 15 chapters of Ataqqinartuaraq (Saint-Exupéry 2017), the terminative case occurs 137 times. 10 of those are cases of conjunction or congruence, and will not be treated here. They will still appear in the appendix.

Below, I will give examples of the semantic roles which the terminative is used for in my corpus. The semantic roles are shown on the semantic map in Appendix I, and the distribution of the semantic roles in Ataqqinartuaraq can be seen in Appendix II.

In Bick 2007, the semantic roles are grouped into different categories. I have been inspired by these groups and adapted the system to Greenlandic. The groups relevant for Greenlandic are: core clause-level roles, spatial and temporal roles, and "other" roles.

9

Liv Molich



Core clause-level rolesThe core clause-level roles cover different kinds of subjects and objects. However, as the terminative case is never in agreement with the verb, all terminatives are more or less loosely boundto the verb. I will here look into which kinds of objects are covered by the terminative.

§REC

The recipient is a +HUM noun which receives something without benefiting from it.

(G5) Kunngi-mut oqar-po-qKing.TRM.s say.IND.3s@ADVL @PRED§REC §PRED

((X.85) (...) he said to the king.)

(G6) Ili-nnut malit-tari-sa-ssa-t qanoq ittuu-ga-mik?You.TRM follow.ITE.PART.ABS.p what be.CAU.3p.COREF@ADVL @SUBJ @ADVL @PRED§REC §TH §ATR §PRED

((XIV.18) What are the orders that you have received?)

Example (G5) shows a simple sentence, where the TRM is adverbial to an intransitive verb. Example (G6), however, is more complicated, but not unusual in Greenlandic. Here, the terminativeis used for the recipient of the orders. The verb has been nominalized, and thus ilinnut is an adverbial to the inderived, transitive verbal stem malig-. From this can be seen that when a transitive verb is nominalized, the terminative flag marks the noun which would have been subject for the transitive verb.

The §REC role only occurs three times. Only in example (G5), a Danish preposition is used: til.

§COG

The cognizer role is used for a +HUM noun who perceives something, but neither experiences it norbenefits from it. The cognizer is often a personal pronoun, like example (G7).

(G7) Taamaammat aatsaar-neq ua-nnut qaqutigoortu-u-vo-qTherefore jawn.doing.ABS.s me.TRM curios.be.IND.3s@CL-ADVL @SUBJ @ADVL @PRED§ADVL §TH §COG §PRED

((X.20) Therefore, yawning is a curiosity to me.)

10

Liv Molich



In five of the seven examples in my corpus, the cognizer is marked with the preposition for in the Danish translation. In the last two occurrences, no preposition is used.

§EXP

The experiencer regularly cannot be distinguished from the cognizer, because the choice of role depends on the reader’s interpretation of the text. Both are +HUM. For example, in IV.47, does ilisimaare-, be familiar with, tell us something about your experiences or just something that you have heard? This could be an argument for combining them into one role. However, an argument for keeping the roles separate could be that all cognizers are subjects of intransitive verbs, while all safe experiencers in the corpus are adverbial to verbs, which are inderived with a causative affix, either {nar} or {tit}.

In (G8), {nar} changes the subject of the inderived verb, qasu-, to be tired, into an oblique object, while the new subject of the clause becomes the causer.

(G8) Nassuiaassi-uaannar-tariaqar-ner-lu meeqqa-nut qasu-nar-ta-qa-a-qExplain.always.need-for.doing.ABS.s.and child.TRM.p tired-be.CAUS.ITE.much.IND.3s@SUBJ @ADVL @PRED§TH §EXP §PRED

((I.21) (…) and it is tiresome for children to have to explain things always.)

In (G9), a transitive verb is inderived with {tit}, which changes it to a ditransitive verb. That changes the original subject into an oblique object in terminative case like in (G8), but in (G9), the new subject corresponds to the original object of the inderived verb.

(G9) Suliar-a inersimasu-nut taku-til-lugu Work.1sPOSS.ABS.s adults.TRM.p see.CAUS.CONT.3sO@OBJ @ADVL @ADVL§TH §EXP §ADVL

((I.10) I showed my work to the adults (…))

In (G10), an intransitive verb is inderived with {tit}, but half-transitivized afterwards. This makes the original subject turn into a terminative adverb, and a new subject is added.

(G10) Taamar-suaq pissaane-qar-neq ataqqinartu-aqqa-mut Such.much power.have.doing.ABS.s prince.little.TRM.s@ADVL @SUBJ @ADVL§ADVL §AG §EXP

11

Liv Molich



tupigutsat-sit-se-qa-a-qwonder.CAUS.HTR.very.IND.3s@PRED§PRED

((X.56) Having such much power made the Little Prince wonder)

The Danish prepositions are different for all three occurences. This is due to different choices of syntactic structures in the Greenlandic and Danish sentences.

§BEN

Like the previous three roles, the beneficiary role is used for nouns which are +HUM or owned by a+HUM. This is often shown by a possessum marking on the noun, and not necessarily with a personal pronoun.

(G11) Illil-li ulloriaati-n-nut iluaqutaa-nngi-la-titYou.but star.2sPOSS.TRM.p advantage-be.NEG.CONT.2s@SUBJ @ADVL @PRED§AG §BEN §PRED

((XIII.106) But you are of no use to your stars…)

The Danish preposition for corresponds to the beneficiary, when marked.

§TH

The semantic role of a theme is a broad category, and is a -HUM object. It can be used for both monocongruent and dicongruent verbs.

(G12) Ataqqinartu-araq naasu-mut qula-ler-sima-vo-qPrince.little.ABS.s flower.TRM.s doubt.INCH.PERF.IND.3s@SUBJ @ADVL @PRED§COG §TH §PRED

((VIII.50) (…) the Little Prince had come to doubt the flower.)

In (G12), a transitive, monocongruent verb takes a terminative object. In (G13), the terminative caseis caused by the incorporation of the noun inissaq, space. This can be seen from the syntactic functions, which – occasionally – can be adnominal and not adverbial. Incorporations like these usually take an instrumental adnominal, but the use of terminative here can have been influenced bythe Danish preposition til in the corresponding Danish sentence.

12

Liv Molich



(G13) Aqqusern-up qulli-a-nut inissa-qa-qqatar-po-qStreet.REL.s lamp.3sPOSS.TRM.s space.have.just.IND.3s@POSS @ADN @PRED§POSS §TH §PRED

((XIV.3) There was just enough room for a street lamp (…))

In example (G14), the terminative is used for the thing that the friends are being warned against. When changing the syntax, the terminative adverbs in all examples so far have been turnable into a subject or an object. This cannot be done for this sentence. Hence, this terminative must be more loosely bound to the verb.

(G14) Titartagaq peqqissaarute-qqissaar-pa-ra Drawing.ABS.s careful-with.careful.IND.1s.3sO@OBJ @PRED§TP §PRED

ikinnguti-kka mianerso-qqu-llu-git navianartu-mutfriend.1sPOSS.p care-take.CAUS.CONT.3pO.COREF danger.TRM.s@OBJ @ADVL @ADVL§BEN §ADVL §TP

((V.54) I have been so careful with my drawing in order to warn my friends against a danger (…),

The themes have different prepositions in the Danish translation. This is partly due to different syntactic structures, e.g. that some of the adverbs are more loosely bound to the verb. This – the loosely bound terminatives – could be an argument for treating the §TP in another group than that of the core clause-level roles.

§REFL

A noun showing that a verb is reflexive or reciprocal is marked §REFL. In the roles used for Portuguese (Bick 2007), §REFL has been put into a category called “Purely syntactically motivated "dummy" roles”. I, however, have chosen to include it in the “Core clause-level roles”. I will not dig deeper into the discussion of where it should be placed, but just note that the §REFL is used for an action against oneself, oneselves or each other in Greenlandic. Thus, it has the role as a subject as well as an object, and thereby should be treated as a core clause-level argument in my definition. Example (G15) does not differ in syntactical structure from (G5).

13

Liv Molich



(G15) (…) imminut oqarfigivoq.Self.TRM.s say-to.IND.3s@REFL @PRED§REFL §PRED

(VII.63 (…) he said to himself)

§REFL is highly frequent in Ataqqinartuaraq, but this might be due to the genre and may not be so frequent in other corpora.

The reflexive in Greenlandic corresponds to reflexive pronouns in Danish, often with the Danish preposition til.

Spatial and temporal rolesThe semantic role of destination or direction is a widely used function of the terminative case. This is not at all surprising, as it is also called allative, which expresses a direction towards something. This disagreement about the designation of the case indicates that it is an example of case syncretism.

Also, space and tempus often share case. This is also true for Greenlandic. Temporal roles for Greenlandic are §LOC-TMP, §ORI-TMP, §DES-TMP, §EXT-TMP, and §FREQ. As the temporal function is not very frequent in Ataqqinartuaraq, I will focus on the spatial roles here.

§DES

§DES is the marking of a destination or a direction. Taking a closer look on §DES, it is clear that it can be subdivided into a number of functions. This will be an obvious thing to do, as the §DES is expressed by lots of different prepositions in Danish. With subdivisions, it might be easier to get a grasp on which preposition to use when.

The first will be §DES-GOAL, which is the movement towards a destination ending at this destination.

(G16) Generali pakkalussa-tut naasu-miit naasumut General.ABS.s butterfly.AEQ.p flower.ABL.s flower.TRM.s@SUBJ @ADVL @ADVL @ADVL§AG §COMP §ORI §DES-GOAL

timmeralaa-qqu-suu-gu-k-kufly.CAUS.someone-be.CAU.1s.3sO@PRED§PRED

14

Liv Molich



((X.62) If I ordered a general to fly from one flower to another like a butterfly (…))

A further subdivision of this will be §DES-INTO which is a place not only reached but also moved into.

(G17) Innuttam-mi pe-qqu-gu-k-kit People.ABS.p.but do.CAUS.CAU.2s.3pO@OBJ @PRED§REC §PRED

imma-mut immi-nnut igissasutsea.TRM.s self.REFL.p throw.FUT.PAR.3p@ADVL @REFL @CL-CIT§DES-INTO §REFL §ACT

((X.67) If you ordered the people to go and throw themselves into the sea (…))

This is only altered a little with the next sub-role, §DES-ONTO, which is basically the same, but with movement onto and not into a destination.

(G18) Ataqqinartu-araq nerrivim-mut qaqi-vo-qPrince.little.ABS.s table.TRM.s climb.IND.3s@SUBJ @ADVL @PRED§AG §DES-ONTO §PRED

((XV.6) The Little Prince climbed onto the table (...))

The next subdivision will be §DES-LOC, which is a direction towards a destination, but without movement towards it.

(G19) Ataqqinartu-aqqa-l-li eqqarsaati-n-nit alla-mut saa-tip-pa-a-ngaPrince.little.REL.s.but thought.my.ABL.p other.TRM.s turn.CAUS.IND.3s.1sO@SUBJ @ADVL @ADVL @PRED§AG §ORI §DES-LOC §PRED

((VII.26) But the Little Prince made me turn away from my thoughts to something else.)

The fourth subdivision is a §DES-DIR, which is a direction towards a destination, which cannot be reached.

(G20) Kunngi-p (...) assa-ni siaar-pa-a (...) King.REL.s hand.3sPOSS.COREF.ABS.s extend.IND.3s.3sO@SUBJ @OBJ @PRED§AG §PAT §PRED

15

Liv Molich



ullorissa-t (...) tunga-an-nut star.REL.p towards.3pPOSS.TRM

@POSS @ADVL§POSS §DES-DIR

((X.47) The king made a gesture with his hand towards the stars (...))

The fifth subdivision will be §DES-SEM, which is a locative-like function, or at least a lexicalized function caused by the semantics of the verb. Visiting someone does not directly say anything aboutmovement; but of course, you must move to visit someone.

(G21) Manna kunngi-mut pulaarni-n-nit nuanner-neru-ngaar-a-miThis king.TRM.s visiting.1sPOSS.ABL.s nice.COMP.much.CAU.3p.COREF@SUBJ @ADVL @ADVL @PRED§TH §DES-SEM §COMP §PRED

((XI.17) This is more entertaining than my visit to the king (...))

Even with this sub-division of the §DES-roles, the picture drawn for the Danish prepositions is still not fully distinct:

§DES-GOAL hen (adverb), op på, til, til, her (adverb), ned her (adverb), til

§DES-INTO i, ud i, ud i, ud i, ned i, ud i, ud i

§DES-ONTO op på

§DES-LOC ud af, over

§DES-DIR imod, om, mod

§DES-SEM hos, til, i, ned (adverb)

This might be due to the different ways of expressing motion towards a place in Greenlandic and Danish. I Danish, a combination of adverb and preposition is often used, while the Greenlandic verbor the nominal root tunge (G21) shows the movement, and the possessor or adverb the direction. In a more detailed study, the syntactic difference should play a major role, as my subdivision here clearly does not help the translations much.

§LOC, §PATH and §EXT

Other spatial roles are location, path and extension, but they occur too seldom in my corpus to be usable for this investigation.

16

Liv Molich



Other rolesCauses and results are verbal roles in Eckhard Bick’s system. Due to the many inderivations and nominalizations in Greenlandic, I use the verbal role §FIN and §CAU for nouns with a verbal content.

§FIN

The §FIN is a purpose, or the reason why you are doing something. It is quite frequent in the corpusdue to many comments on “having time for” and “what use something is for”. This might be different in another corpus.

(G22) Sinnattuaqini-ssa-n-nut piffissa-qa-nngi-la-ngaUtopian-dream.FUT.1sPOSS.TRM.s time.have.NEG.IND.1s@ADVL @PRED§FIN §PRED

((XIII.43) I don’t have time for utopian dreaming.)

§CAU

The causative do have much in common with the purpose, and can be hard to distinguish between. While the §FIN explains what you are doing something for, the §CAU explains what is causing youto do it.

(G23) Ataqqinartu-aqqa-p paasi-sinnaa-nngi-la-a aqqusernu-p Prince.little.REL.s understand.able-be.NEG.CONT.3s.3sO street.REL.s@SUBJ @PRED @POSS§AG §PRED §POSS

qulli-a quller-mil-lu ikitsissut su-mut iluaqutaa-ner-su-tlamp.3sPOSS.ABS.s lamp.INS.s.and lighter.ABS.s what.TRM advantage-be.if.PAR.3p@SUBJ @ADN @SUBJ @ADVL @ADVL§TH §TH (§TH) §CAU §TH

((XIV.4) The Little Prince didn’t understand what a street lamp and a lamplighter could be an advantage for [why he did it])

§SOA and §TP

The §SOA and §TP are specifications of the §TH. The §SOA is a state of affairs, a proposition or fact, while the §TP is a topic of an action. In Bick’s grouping, they are placed within the “Core clause-level roles”. However, for Greenlandic, they should be treated in the same way as §FIN and

17

Liv Molich



§CAU, since none of them are valency arguments even though they are closely connected to the verb. This also accounts for some of the §TH’s, as mentioned above. The frequency in the corpus is too low to say any more about this category in this paper. For none of them, a single preposition cancover the translations, so a closer look on the roles is needed.

§ATR, §CONT and §COMP

The attributive and the content are both placed in the “Attributive and other "close" roles” category in Bick 2007, while the comparative is placed in the “Adverbial roles” together with §FIN and §CAU. My corpus does not provide the basis for retaining this distinction. A later study with more focus on the dependencies might reconsider this.

The comparative is often marked by AEQ or {niru} in combination with ABL. Here, it is used for {vallaar} in combination with TRM.

(G.24) Miki-vallaa-rujus-suar-pu-gut inner-mik anitsisartu-uti-ttaSmall.too.much.very.IND.3p fire.INS.s vulcano.owned.1pPOSS.REL.p@PRED @ADVL @POSS§PRED §INS §POSS

saner-tar-ni-ssa-an-nut.Sweep.ITE.doing.FUT.3pPOSS.TRM@ADVL§COMP

((IX.12) (…) we are far too small to sweep our vulcanos.

The use here is thus lexical, and in this way closely linked to the verb. The same can be said about the §CONT, which is triggered by inissaqar-, have space (for), while the §ATR can be used for comparisons as well as §COMP. They are not frequent enough for further discussion.

Lexicalized usageIn the roles above, I have mentioned the term “lexicalized” a few times. When using this term, I refer to the combination of the verb and the noun, which in the terminology of Langgård and Langgård (1988: 156-157) is called “construction” rather that “rection”. The construction is lexicalized, when it is frequent, or when the semantics is not longer clear. Apart from the examples mentioned above, one construction is relatively frequent in Ataqqinartuaraq: ilumut, true, which fundamentally means towards the inner (part), and derivations hereof. Also case shift for constructions with certain semantics can be lexicalized, like we saw in (G3) and (G4).

18

Liv Molich



DiscussionWe have seen how the terminative case is used in Greenlandic, which semantic roles are covered bythe terminative case and which structures are most frequent as regards some of the roles. The resultsfrom these investigations can be seen on the semantic map in Appendix I, and the number of occurrences of each semantic role can be easily perused in Appendix II. How the different semantic roles are translated in the Danish version of the corpus can be seen from both appendices.

As we saw in the review of the semantic rules above, some of the semantic features seem to be closely related to syntax. Therefore we should build a more precise syntactic analysis into the syntactic part of the machine translation project where now all terminatives are adverbials, adnominals, or reflexives. This improvement could be accomplished by benefitting from transformational grammar theory, or by carefully marking the dependencies and their correlation to the semantic roles.

From the semantic map and the distributional table, a somewhat clear distribution of the most central roles and their translations can be seen. However, the conclusions are based on only one text, and this text is not originally written in Greenlandic. Therefore, a more thorough analysis of the terminative case in Greenlandic should include a bigger and more varied corpus, consisting of original Greenlandic text of differing complexities, by different authors, and in different styles. Also, the map must not stand alone, since the method is quantitative and only qualitative to a low degree. The map should be a method for reviewing and comparing languages and for being inspiredto change things in the description of the languages.

In Greenlandic, the subject and/or object is marked in the flexion, and it is often not marked explicitly. Hence, often neither subject and objects, nor other arguments will be found. This makes the need for a bigger corpus even more clear. A study of when the core clause-level terminatives areexplicitly expressed would be interesting when the reverse translation project from Danish to Greenlandic begins.

ConclusionIn this paper, I have looked at the multifunctionality of the West Greenlandic terminative case investigating a corpus containing 137 terminatives.

I have shown how the terminatives can be mapped semantically, and that some of the semantic rolesshare syntactic features. The distribution of terminatives is visually shown on a semantic map and ina table, and this shows that the terminatives are approximately evenly distributed between “core syntactic”, “spatial and temporal” and “other” roles. The map shows a correlation between the semantic roles and the Danish prepositions til and for, but across the three groups.

19

Liv Molich



However, the results are not completely reliable as the corpus is relatively small and unvaried. An investigation of a larger corpus and an expansion of the project to include the other cases as well would be useful.

20

Liv Molich



Literature

TheoryÁgel, Vilmos, and Klaus Fischer (2015 [2010]): “Dependency Grammar and Valency Theory”. In: Bernd Heine and Heiko Narrog (eds.): The Oxford Handbook of Linguistic Analysis. New York: Oxford University Press, 225-257.

Bick, Eckhard (2007): Semantic Roles for Automatic Corpus Annotation, http://beta.visl.sdu.dk/~eckhard/pdf/semantic_roles_manual.pdf.

Bick, Eckhard (2011): A FrameNet for Danish, http://visl.sdu.dk/~eckhard/pdf/nodalida2011_framenet.pdf.

Fillmore, Charles (1968): “The Case for Case”. In: E. Bach & R. Harms (eds.): Universals in linguistic theory. New York: Holt, Rinehart and Winston, 1-88.

Fortescue, Michael (1984): West Greenlandic. London, Sydney and Dover, Newhampshire: Croom Helm.

Hansen, Erik, and Lars Heltoft (2011): Grammatik over det Danske Sprog. Det Danske Sprog- og Litteraturselskab.

Haspelmath, Martin (2003): “The Geometry of Grammatical Meaning: Semantic Maps and Cross-Linguistic Comparison”. In: Tomasello, Michael (ed.): The new psychology of language, vol. 2. Mahwah, New Jersey and London: Erlbaum, 211-242.

Jurafsky, Daniel, and James H. Martin (2015): “Semantic Role Labeling”. In: Speech and LanguagePossessing. 3rd edition (draft), https://web.stanford.edu/~jurafsky/slp3/22.pdf

Langgård, Karen, and Per Langgård (1988): Oqaasilerinermik Ilinniutit. Elementær Grønlandsk Grammatik. Manuscript. Nuuk: Ilisimatusarfik.

Langgård, Karen (1992): Øvebog i Grammatik. Nuuk: Ilisimatusarfik.

Malchukov, Andrej, and Heiko Narrog (2009): “Case Polysemy”. In: Andrej Malchukov and Andrew Spencer (eds.): The Oxford Handbook of Case. New York: Oxford University Press, 518-534.

Nielsen, Flemming A.J. (2014): Vestgrønlandsk Grammatik. Manuscript, http://www.groenlandskgrammatik.dk/Groenlandsk_grammatik_web.pdf.

Næss, Åshild (2009): “Varieties of Dative”. In: Andrej Malchukov and Andrew Spencer (eds.): The Oxford Handbook of Case. New York: Oxford University Press, 572-580.

21

Liv Molich



Sadock, Jerrold M. (2003): A Grammar of Kalaallisut (West Greenlandic Inuttut). Languages of the World/Materials 162. München: LINCOM.

Tesnière, Lucien (1976): Éléments de Syntaxe Structurale. Paris: Klincksieck.

Zúñiga, Fernando (2011): “Why Should Beneficiaries be Subjects (or Objects)?” In: Seppo Kittilä, Katja Västi and Jussi Ylikoski (eds.): Case, Animacy and Semantic Roles (Typological Studies in Language 99), 329–348. Amsterdam: Benjamins.

CorporaKalaallit Nunaata Radioa (2005): Aallaasiutillit nakkutiginninnerusariaqarput, https://knr.gl/kl/nutaarsiassat/aallaasiutillit-nakkutiginninnerusariaqarput, September 13th.

Kalaallit Nunaata Radioa (2007): Ataataasoq pillagaasussanngorpoq, https://knr.gl/kl/nutaarsiassat/ataataasoq-pillagaasussanngorpoq, June 12th.

Saint-Exupéry, Antoine de (1943): The Little Prince. Translated by Katherine Woods from French Le Petit Prince (1943). New York: Reynal & Hitchcock.

Saint-Exupéry, Antoine de (1991): Den Lille Prins. Translated by Asta Hoff-Jørgensen from French Le Petit Prince (1943). Viborg: Lindhardt og Ringhof.

Saint-Exupéry, Antoine de (2017): Ataqqinartuaraq. Translated by Mira Kleist, with help from Juana Petrussen and Carla Rosing Olsen, from French Le Petit Prince (1943). Nuuk: Milik Publishing.

ToolsThe Language Secretariat of Greenland: Dictionary, https://oqaasileriffik.gl/search/dictionary/. A lexical base for the machine translation project, also serving as a dictionary.

The Language Secretariat of Greenland: Live Analysis, https://oqaasileriffik.gl/langtech/live/. A constraint grammar based tool for sentence analysis.

University of Southern Denmark and GrammarSoft: Danish Framenet Lookup Interface, http://visl.sdu.dk/visl2/framenet_lookup.html. A lookup tool for Danish semantic frames.

22

Liv Molich



Appendix I: Semantic maps on terminative in Greenlandic

__________________| SPATIAL AND ||TEMPORAL ROLES|| || Location (6) | | §LOC | | §PATH || §EXT | __________________________________| | | CORE CLAUSE-LEVEL ROLES || | | || | –– –– | Recipient (25) –– –– Beneficiary (5)|| Direction (26) | | §REC §BEN || §DES | | §REFL || | | || | til | | ||_________________ | | | for |

| | | | | Experiencer (10) | | §COG | Purpose (21) | §EXP | §FIN | | §CAU |__________________________________|

The map shows the semantic functions for the West Greenlandic terminative case. The map is basically the same as the dative map (Haspelmath 2003), except for the predicative possessor, the external possessor and the dative judicantis. The predicative passive and the external possessor are not covered by the Greenlandic terminative. The dative judicantis is covered by the Greenlandic terminative, but the distinction between this and the experiencer is not completely clear. Therefore, the judicantis is primarily contained by the experiencer.

The squares indicate that the terminative case is used for at least two widely different purposes – spatial and temporal roles, as well as core clause-level roles. Outside the squares are the remaining roles, of which I have only mapped §FIN and §CAU. Not all the functions of the terminative are shown on the map.

The blue figure illustrate where the Danish complement is taking the preposition til, while the red figure encircles the Danish preposition for.

23

Liv Molich



Appendix II: Distribution of semantic roles for the Greenlandic terminativeThe distribution of terminatives in Ataqqinartuaraq is shown in the table below. Even though the distribution is calculated from a relatively small number of sentences in a undifferentiated corpus, the distribution diagram still gives an impression of what the terminative case is used for in West Greenlandic. The second row shows the distribution of the semantic roles for the three sub-groups, in total 118. The two bottom rows show the distribution of functions with at least five occurrences, in total 88.

Core clause-level Temp/Spat Other Total

50 = 42,4% 40 ≈ 33,9% 28 ≈ 23,7% 118 = 100%

§COG §BEN §TH §REFL §DES §FIN

7 ≈ 8,0% 5 ≈ 5,7% 10 ≈ 11,5% 22 = 25% 26 ≈ 29,5% 18 ≈ 20,5% 88 = 100%

44 = 50% 26 ≈ 29,5% 18 ≈ 20,5% 88 = 100%

The distribution can be further investigated from the list below. It shows the number of mappings for each semantic role, and how it is translated in the Danish version of the text when prepositions are used.

Total number of terminatives in corpus: 137 (chapters 1-15)

Core clause-level roles (50)

§REC: 3 til (1)

§COG: 7 for (4)

§EXP: 3 for (1), med (1)

§BEN: 5 for (3)

§TH:10 over (3), mod (2), på (1), til (1)

§REFL: 22 til (9), ad (1), over (1), over for (1)

Spatial roles (32)

§DES: 26

§DES-LOC ud af (1), over (1)

§DES-GOAL til (3), op på (1)

24

Liv Molich



§DES-INTO ud i (5), ned i (1), i (1)

§DES-ONTO op på (1)

§DES-DIR imod (1) om (1), mod (1)

§DES-SEM hos (1), til (1), I (1)

§LOC: 3 i (1) om (1), på (1)

§PATH: 1 gennem (1)

§EXT: 2 fra (2)

Temporal roles (8)

§LOC-TMP: 1 i (1)

§ORI-TMP: 2 til (1)

§DES-TMP: 2 indtil (1)

§EXT-TMP: 1 til (1)

§FREQ: 2 i (1)

Other roles (28)

§FIN: 18 til (6), med (4), i (1), ud af (1)

§CAU: 3 mod (1)

§SOA: 2 for (1), på (1)

§TP: 2 om (1)

§ATR: 1

§CONT: 1 til (1)

§COMP: 2 til (1)

Other terminatives: 19

Congruence (8)

Conjunction (2)

Lexicalization: TRM tunnga (2)

Lexicalization: ilumut (7)

25

Liv Molich



Appendix III: Semantic roles and other abbreviationsFor more information about usage, see Langgård and Langgård 1988 as regards Greenlandic grammar, and Bick 2007 as regards semantic roles.

Syntactic functions@ADN = adnominal verb or noun

@ADVL = adverbial verb or noun

@AG = agent

@CL-ADVL = clause adverbial

@CL-CIT = the main verb in an oratio oblique clause

@N-CONG = adnominal noun in congruence

@OBJ = object

@POSS = possessor

@PRED = predicative

@SUBJ = subject

Semantic roles§ACT = action

§ADVL = adverbial verb, noun or particle

§AG = agent

§ATR = attributive

§BEN = benefactive

§CAU = cause

§COG = cognizer

§COMP = comparation

§CONT = content

§DES = destination and direction towards a goal

26

Liv Molich



Subdivisions: §DES-GOAL = movement towards a destination, which is reached

§DES-INTO = movement towards and into a destination

§DES-INTO = movement towards and onto a destination

§DES-LOC = movement on a location, facing a destination

§DES-DIR = movement towards a destination, which cannot be reached

§DES-SEM = locative-like function, or at least a lexicalized function caused by the semantics of the verb

§EXP = experiencer

§EXT = extension

§FIN = purpose

§INS = instrument

§LOC = location

§ORI = origin and direction away from

§PATH = path

§PAT = patient

§POSS = possessor

§PRED = predicative

§REC = recipient

§REFL = reflexive or reciprocal

§SOA = state of affairs, proposition, or fact

§TH = theme

(§TH) = theme in congruence with another theme

§TP = topic

Other abbreviationsABL = ablative case

ABS = absolutive case

27

Liv Molich



ACT = active voice

AEQ = aequative case

CAU = causative mood

CAUS = causative affix

COMP = comparative affix

CONT = contemporary mood

COREF = co-referential inflexion

FUT = future tense affix

HTR = “half-transitive” affix

HUM = human, either + or -

INCH = inchoative affix

INDEF = indefinite article

IND = indicative mood

INS = instrumental case

ITE = iterative affix

NEG = negating affix

p = plural subject or noun

pO = plural object

pPOSS = possessed by a plural noun

PART = passive participle

PASS = passive voice or passivizing affix

PERF = perfective aspect affix

PRP = preposition

REL = relative case

s = singular subject or noun

sO = singular object

28

Liv Molich



sPOSS = possessed by a singular noun

PAR = participle mood

TRM = terminative case, also called allative

29

Liv Molich



Appendix IV: CorpusBelow, all occurrences of terminative case in Ataqqinartuaraq are listed, the terminatives in bold font. The roman numbers correspond to the chapters, while the arabic numbers correspond to the sentence number.

Greenlandic Danish ROLEI (Dan.I.1)Arfinilinnik ukioqarlunga assilissamik tupinnartumik takuvunga kiattup orpippassuipillugit atuakkiami, imatut atilimmi ”Oqaluttuat ilumoortut”.

Da jeg var seks år, så jeg engang et pragtfuldt billede i en bog om urskoven, der hed Sandfærdige Historier.

(ilumut)

(Dan.I.10)Suliara inersimasunut takutillugu aperaakkataanna ersigineraat.

Jeg viste de voksne mit mesterværk og spurgte, om de ikke blev bange.

§EXP

(Dan.I.21)Inersimasut nammineerlutik imaaliallaannaq paasinnissinnaanngillat, nassuiaassiuaannartariaqarnerlu meeqqanut qasunartaqaaq.

De voksne forstår aldrig noget helt af sig selv, og det er så trættende for børn evig og altid at skulle komme med forklaringer.

§EXP

(Dan.I.27)

Inuunera tamaat inuit ilumoorsaartut susassaqarfigisimaqaakka.

Jeg har således hele mit liv haft en masse at gøre med en masse alvorlige mennesker.

(ilumut)

II (Dan.II.1)Tassa kisima inuusimavunga, iluamik paaseqatigisinnaasannik oqaluussassaqarnanga, ukiut arfinillit matuma siorna Saharami maskiinaajarnissama tungaanut.

Jeg har altså levet alene, uden et menneske jeg rigtig kunne tale med, lige indtil for seks år siden, da jeg fik maskinskade i Sahara-ørkenen.

§DES-TMP

(Dan.II.4)Tamanna uannut inuunermik akeqarsinnaavoq.

Det betød liv eller død for mig. §COG

(Dan.II.5)Sapaatip-akunnerinnaanulluunniit imissaranaammanngilaq.

Jeg havde knap nok drikkevand til en uge.

§EXT-TMP

(Dan.II.6)Tassalu unnuk siulleq sioqqanut innarpunga inoqarfimmut qaninnermut kilometerinik tuusintilinnik ungasitsigisumi.

Den første aften lagde jeg mig da til atsove i sandet tusinde mile borte fra beboede egne.

§LOC

§EXT(Dan.II.24)

30

Liv Molich



Meeqqatut isikkoqanngivippoq inoqajuitsumiittutut tammarsimasutut inoqarfimmut qaninnermut kilometerinik tuusintilinnik ungasitsigisumiittutut.

Han lignede på ingen måde et barn, der var blevet borte midt i ørkenen, tusinde mile fra beboede egne.

§EXT

III (Dan.III.44)

- Arlaannut iluarisaminut. Hvor skulle det dog gå hen?§DES-GOAL +congruence

(Dan.III.45)Hvor det fandt for godt.

IV (Dan.IV.36-37)

Ataqqinartuaqqap inuusimaneranut uppernarsaatissat tassaapput inequnarluinnarnera, illartarnera savaatitaarusunneralu; savaatitaarusunnermi uppernarsaataavoq inuusimanermut”.

Beviset for at den lille prins har eksisteret er, at han var bedårende, at han lo og at han gerne ville have et får; når man vil have et får, er det da etbevis på, at man eksisterer; så ville de trække på skuldrene og betragte detsom barnesnak.

§SOA

§SOA(Dan.IV.47)

Inuunermik ilisimaarinnittumut tamanna ilumoorpalunnerussagaluaqaaq.

For den, der kender livet, ville det have lydt meget rigtigere.

§COG

(ilumut)(Dan.IV.69)

Ajoraluartumilli uanga sava karsip iluaniittoq akimut takusinnaasimanngisaannarpara.

Men jeg har desværre aldrig kunnet se et får tværs gennem en kasse.

§PATH

(Dan.IV.70)Imaassinnaavormi inersimasunut assingulaartunga.

Måske ligner jeg alligevel de voksne lidt.

§ATR

V (Dan.V.17)- Baobabit aallaqqaammut mikisuusarput - allinnginnerminni.

Baobaberne har været små til at begynde med - inden de blev store.

§ORI-TMP

(Dan.V.27-28)Naatsiiassalli ersinngitsuupput nunami toqqorlutik sinittartut, ataatsip iterusulernissaata tungaanut.

Men frøene er usynlige, de ligger og sover i jordens skjul, indtil det passer et af dem at vågne op.

§DES-TMP

(Dan.V.29)Taava tasitsaartarpoq, aallaqqaammullu qunusunnguamik naanera ataasinnguaq inequnavissoq ajoqutaanngivilluni seqernup tungaanut siaartarpoq.

Så strækker det på sig og skyder førstfrygtsomt en lille yndig og harmløs spire op imod solen.

§ORI-TMP

31

Liv Molich



§DES-DIR(Dan.V.39)

Ullaakkut imminut isumagereeraanni ulloriaaraq isumageqqissaarneqartariaqarpoq.

Når man er færdig med at gøre sig selv i stand om morgenen, så må man omhyggeligt gøre planeten i stand.

§REFL

(Dan.V.50)

Baobabinulli tunngatillugu tammartajaarluni ulloriaaqqamut pigaanni, tamatuma kingunissai ajortut ilisimanngisat annertoqimmata, tamatumuuna pinaasernera akiorpara.

Men risikoen ved baobaberne er så lidet kendt, at de farer man udsætter sig for, når man forvilder sig op på en asteroide, er så store, at jeg for een gangs skyld vil overvinde denne modvilje.

(tunnga)

§DES-GOAL(Dan.V.52-53)

”Meeqqat, baobabinut mianersoritsi!” Børn, tag jer i agt for baobaberne! §CAU(Dan.V.54)

Titartagaq peqqissaaruteqqissaarpara ikinngutikka mianersoqqullugit navianartumut nalusatsinnut qangali uatsinnik navianartorsiortitsisimasumut.

Det er for at advare mine venner mod en fare, som både de og jeg længe har været udsat for uden at vide det,at jeg har gjort mig sådan umage med den tegning.

§TH + congruense + congruense

VI (Dan.VI.2B)Inuunerit alianartoq piffissap ingerlanerani uannut paasinarsiartuaarpoq.

Lidt efter lidt kom din bedrøvelige tilværelse til at stå klart for mig.

§COG

(Dan.VI.10-11)Tupaallappaseqqaarputit, taavali imminut illaatigiitigalutit oqarputit:

Da så du først meget overrasket ud, men så lo du ad dig selv og sagde:

§REFL

(Dan.VI.15)Minutsip ataatsip ingerlanerani Frankrigimut angalapallassinnaasuugaanni seqernup tarrinnera takusarneqarsinnaagaluarpoq.

Hvis man så bare kunne rejse til Frankrig på eet minut, kunne man nå at få solnedgangen at se.

§DES-GOAL

VII (Dan.VII.8)- Taavami kapinartut sumut iluaqutaappat? Men hvad nytter så tornene da? §FIN

(Dan.VII.15A)- Kapinartut sumulluunniit iluaqutaanngivipput.

Torne er ikke til nogensomhelst nytte.§FIN

(Dan.VII.21)Sapinngisaminnik imminnut paarisarput. De sikrer sig så godt de kan. §REFL

(Dan.VII.26)Ataqqinartuaqqalli eqqarsaatinnit allamut saatippaanga.

Men den lille prins rev mig ud af mine betragtninger.

§DES-LOC

32

Liv Molich



(Dan.VII.35)

Nakkuppaanga kaataq tigullugu nikorfasunga assakka tarnummik qernertuinnaasut, suunersumut kusanaagingaagaanut sikingallunga.

Han stirrede på mig, som jeg stod der med en hammer i hånden, hænderne sorte af smørelse, bøjet over en genstand, som han syntes var forfærdelig grim.

§DES-LOC + congruence

(Dan.VII.41)

Ileqimisaartorpoq nutsani imusut kuultitulli sungaartigisut sumut tamaanga erfatsillugit.

Han rystede på hovedet, så de guldgule krøller stod om hovedet på ham.

§DES-DIR + congruence

(Dan.VII.64)”Pavani arlaanniippoq uanga naasuutiga ...” imminut oqarfigivoq.

Et eller andet sted deroppe findes min blomst...(Dan.VII.63)siger han til sig selv. §REFL

VIII (Dan.VIII.7)

Ataqqinartuaqqap malinnaaffigaa sikkersussarujussuaq alliartortoq, qularinagulu ilimasungaalerpoq tupinnartumik inerittoqarumaartoq, naasulli kusatsinnissaminut piareersarnini naammassinaveerpaa taqqamani ineeqqaminiqorsooqqissumi.

Den lille prins, der overværede hvordan en kæmpestor knop lejrede sig, havde en meget stærk fornemmelse af, at en forunderlig åbenbaring ville folde sig ud af den, men blomsten kunne ikke blive færdigmed sine forberedelser til at blive smuk, derinde i sit grønne lønkammer.

§FIN

(Dan.VIII.16)Ullaalli ilaanni, seqineq nuilersorlu, naasoq imminut nittarpoq.

Men en morgen, netop da solen var ved at stå op, lod blomster sig se.

§REFL

(Dan.VIII.37)

”Sueq nuannariivinngilaa ... nunamut taamatunngatigisumi suingaarsinnaanavianngitsoq ataqqinartuaqqap oqaatigisimavaa.

Ikke fordrage træk... der kunne virkelig ikke være tale om træk så langt nede ved jorden, havde den lilleprins bemærket.

(tunnga)

(Dan.VIII.43)Tassungali killillunilu imminut unitsippoq. Men her afbrød den sig selv. §REFL

(Dan.VIII.46)

Taamak salluliani sianiilikaartigisoq pillugu paasisaagami kanngusulluni quersutsiarpoq, kanngusuutissaq ataqqinartuaqqamut tutsinniarlugulusooq:

Skamfuld over at lade sig gribe i en så enfoldig løgn havde den hostet et par gange for at få det til at se ud, som omdet var den lille prins, der burde skamme sig:

§DES-GOAL

(Dan.VIII.50)Taamaalilluni ataqqinartuaraq asannikkaluaqaluni naasumut

Således var den lille prins trods sin redebonne kærlighed snart begyndt at

§TH

33

Liv Molich



qulalersimavoq. tvivle på blomsten.(Dan.VIII.62-63)

Naasut imminnut assortortaqigamik, uangaliinuusuppallaarsimavunga asanninnera paasissallugu.

Blomster er så selvmodsigende, men jeg var for ung til at forstå at elske den.

§REFL

IX (Dan.IX.12)Uagulli maani nunarsuatsinni mikivallaarujussuarpugut innermik anitsisartuutitta sanertarnissaannut.

Nu er vi naturligvis alt for små her på jorden til at feje vore vulkaner.

§COMP

X (Dan.X.9)Ataqqinartuaqqap ilisimanngilaa kunnginut silarsuaq nalorninaalluinnartuummat.

Han vidste ikke, at for konger er verden uhyre ligetil.

§COG

(Dan.X.20)Taamaammat aatsaarneq uannut qaqutigoortuuvoq.

Jeg betragter gaben som en kuriositet.

§COG

(Dan.X.30)

Kunngimummi pingaarutilerujussuuvoq oqaasissaqarnerata ataqqineqarnissaa.

Thi det spillede en kolossal rolle for kongen, at hans autoritet blev respekteret.

§BEN

(Dan.X.45)- Sunut tamanut, kunngi akivoq taamaattussaaginnartutut isiginnippasilluni.

Over alt, svarede kongen med den største selvfølgelighed.

§TH + congruence

(Dan.X.46)

- Sunut tamanut? Over alt?§TH + congruence

(Dan.X.47)Kunngip innimisuppaseqaluni assani siaarpaa ulloriaaqqap, ullorissat angalasut ullorissallu allat tungaannut.

Kongen slog i al beskedenhed ud med hånden mod planeten, de andre planeter og stjernerne.

§DES-DIR

(Dan.X.50)

- Tamakkorpassuarnut... kunngi akivoq.Over det alt sammen... svarede kongen.

§TH

(Dan.X.51)Kisimiivillunimi naalakkersuisuuginnanngilaq, silarsuarmullitamarmut kisimiilluni naalakkersuisuuvoq.

Thi ikke alene var han absolut enehersker, han var universel enehersker.

§TH + congruence

(Dan.X.56)Taamarsuaq pissaaneqarneq ataqqinartuaqqamut tupigutsatsitseqaaq.

En sådan magt slog den lille prins med forbløffelse.

§EXP

(Dan.X.58)Ulloriaatinnguanilu qimataq maqaasileramiuk sapiiserluni kunngimut

Og da han kom til at længes efter sin lille forladte planet, dristede han sig til

§REC

34

Liv Molich



noqqaalerpoq. at anmode kongen om en gunst.(Dan.X.62)

- Generali pakkalussatut naasumiit naasumut timmeralaaqqusuugukku, imaluunniit isiginnaartitsissummik allaqqullugu nattoralinngoqqulluguluunniit, generalilu naalanngippat, taava kina pisuussuteqassava, taanna uangaluunniit?

Hvis jeg befaler en general at flyve frablomst til blomst som en sommerfugl,eller at skrive et skuespil eller at forvandle sig i en havørn, og generalen ikke udførte ordren, hvem ville da hav skylden, han eller jeg?

§DES-GOAL

(Dan.X.67)Innuttammi peqqugukkit immamut imminnut igissasut taava pikitsitsissapput.

Hvis du befaler dit folk at styrte sig i havet, gør det oprør.

§DES-INTO

§REFL(Dan.X.80)

Pissaaq imaalinerani ... imaalinerani ... unnugu pissaaq, nalunaaqutaq qulingiluanutminutsinik marlunnik sioqqullugu!

Det bliver henad... henad... det bliver iaften, når klokken er hen imod to minutter i otte!

§LOC-TMP

(Dan.X.85)- Maani allamik pisassaaruppunga, kunngimut oqarpoq.

Jeg har ikke mere at gøre her, sagde han til kongen.

§REC

(Dan.X.89)- Sumut ministeri? Hvilken slags minister? §TH

(Dan.X.90)- Hm ... inatsiseqarnermut ministeri! Hm... justitsminister! §TH

(Dan.X.98)- Taavami imminut eqqartuussinnaavutit, kunngi akivoq.

Så kan du dømme dig selv, svarede kongen.

§REFL

(Dan.X.100)Imminut eqqartuunnissaq allanik eqqartuussinermit nalunarnerujussuuvoq.

Det er meget vanskeligere at dømme sig selv end at dømme andre.

§REFL

(Dan.X.101)Imminulli eqqortumik pineqaatississinnaaguit, taava ilisimatuuissuusimassaatit.

Men er du i stand til at fælde den rette dom over dig selv, da er du en virkelig vismand.

§REFL

(Dan.X.102)- Sumiluunniimmi imminut eqqartuussinnaavunga, ataqqinartuaraq oqarpoq.

Jamen jeg kan jo dømme mig selv hvorsomhelst, sagde den lille prins.

§REFL

(Dan.X.108-109)Taanna ilaannikkut toqumut eqqartuuttarsinnaavat.

Den kan du jo dømme til døden en gang imellem.

§FIN

(Dan.X.110)Taamaalilluni illit naapertuilluarnerit taassuma inuuneranut apeqqutaassaaq.

Dens liv ville således afhænge af dit retfærd.

§BEN

(Dan.X.122-123)

35

Liv Molich



Ilumummi inersimasunuku eqqumiitsut, ataqqinartuaraq silarsuarmut ingerlaqqikkami imminut oqarfigaaq.

Hvor er de voksne dog mærkelige, sagde den lille prins til sig selv, da handrog videre ud i verden.

(ilumut)

§DES-INTO§REFL

XI (Dan.XI.10)Uannik qiimmattaatiginnittunut ilassinnissut.

Den er til at hilse med, når man tiljubler mig.

§FIN

(Dan.XI.11)Ajoraluartumilli tamaanga aqqusaartoqanngisaannarami.

Der kommer bare aldrig nogen forbi her, desværre.

§DES-GOAL

(Dan.XI.17)- Manna kunngimut pulaarninnit nuannernerungaarami, ataqqinartuaraq imminut oqarfigaaq.

Dette her er meget morsommere end at være på besøg hos kongen, sagde den lille prins til sig selv.

§DES-SEM

§REFL(Dan.XI.31)

- Nersorpakkit, ataqqinartuaraq oqarpoq tuinilu kiviimillugit, kisianni tamanna ilinnut sumut iluaqutaassagami?

Jeg beundrer dig, sagde den lille prins og trak lidt på skuldrene, men hvilken interesse kan du dog have i det?

§BEN

§FIN(Dan.XI.33-34)

Inersimasut ilumuuissuaq eqqumeeqaat, ataqqinartuaraq silarsuarmut ingerlaqqikkami imminut oqarfigaaq.

De voksne er ganske afgjort meget mærkelige, sagde han til sig selv, da han drog videre ud i verden.

(ilumut)

§DES-INTO§REFL

XII (Dan.XII.17-18)Inersimasut ilumoortumik eqqumeeqaat, ataqqinartuaraq silarsuarmut ingerlaqqikkami imminut oqarfigaaq.

De voksne er bestemt meget mærkelige, sagde han til sig selv, da han drog videre ud i verden.

(ilumut)

§DES-INTO§REFL

XIII (Dan.XIII.11)Ikeqqinnissaanut piffissaqanngilanga. Har ikke tid til at tænde den igen. §FIN

(Dan.XIII.20)Piffissara asuliinnaq oqalunnermut atorneq ajorpara!

Jeg spilder ikke min tid med snak! §FIN

(Dan.XIII.26)Siulleq ukiut toatiivit matuma siorna pivoq, sullineq nerrivinnut suminngaanniit

Første gang - det er nu toogtyve år siden - det var en oldenborre, der

§DES-GOAL

36

Liv Molich



pinerpoq nakkarmat.dumpede ned her, guderne må vide hvorfra.(Dan.XIII.29)

Naammattumik timersorneq ajorama, pisuttuarnissannummi piffissaqannginnama.

Jeg får ingen motion.

(Dan.XIII.30)Jeg har ikke tid til at gå tur. §FIN(Dan.XIII.33)

Sumummaa killittunga:Hvor var det, jeg kom til: femhundrede og een millioner…

§DES-SEM

(Dan.XIII.42)Uangali ilungersortuugama, sinnattuaqinissannut piffissaqanngilanga.

Men jeg er en alvorsmand.

(Dan.XIII.43)Jeg har ikke tid til at drømme. §FIN(Dan.XIII.48)

- Ullorissammi famhunnorujut millionit taava sumut atorniarakkit?

Og hvad skal du med fem hundrede og een millioner stjerner?

§FIN

(Dan.XIII.51)

- Ullorissammi taakku sumut atorniarpigit?Og hvad skal du så med alle de stjerner?

§FIN

(Dan.XIII.52)- Sumut atorniarnerikka? Hvad jeg skal med dem? §FIN

(Dan.XIII.54)- Sumut atornianngilakka. Ikke noget. §FIN

(Dan.XIII.62)- Ullorissanik piginninnerit illit sumut iluaqutigaajuk?

Og hvad har du så ud af at eje stjernerne?

§FIN

(Dan.XIII.64)- Pisuujunerimmi sumut iluaqutigaajuk? Og hvad gavner det dig, at du er rig? §FIN

(Dan.XIII.65)- Ullorissanik allanik pisisinnaalerninnut, allaqassappat.

Så kan jeg købe flere stjerner, hvis der skulle findes nogen.

§FIN

(Dan.XIII.66)

Ataqqinartuaraq imminut oqarfigaaq:Han tænker jo næsten ligesom den fulde mand, sagde den lille prins til sig selv.

§REFL

(Dan.XIII.81-82)- Taamaalluarsinnaavormi, ataqqinartuaraq oqarpoq, kisiannimi taava sumut atortarpigit?

Det kan være rigtig nok, sagde den lille prins, men hvad bruger du dem til?

§FIN

(Dan.XIII.88)- Kisiannimi soorlu uanga Jamen, hvis jeg ejer et tørklæde, så §LOC

37

Liv Molich



qungasequteqaruma, taava qungatsinnut atitillugu nassarsinnaavara.

kan jeg tage det om halsen og tage detmed mig.(Dan.XIII.91)

- Ilumut, kisianni aningaaserivimmut ilisinnaavakka.

Nej, men jeg kan sætte dem i banken. (ilumut)

§DES-SEM(Dan.XIII.93)

- Ulloriaatima amerlassusii pappiliamut allariarlugit taanna pappiliaq amusartumut ileriarlugu parnaarutissara.

Det betyder, at jeg skriver op på et stykke papir, hvor mange stjerner jeg har.

§LOC

(Dan.XIII.94)Og så låser jeg det stykke papir ned i en skuffe.

§DES-INTO

(Dan.XIII.105)Taamaalillunga taakkuninnga piginnittuunerainnermik anitsisartuutinnut naasuutinnullu iluaqutaavoq.

Det er til gavn for mine vulkaner og det er til gavn for min blomst, at jeg ejer dem.

§BEN + conjunction

(Dan.XIII.106)

Illilli ulloriaatinnut iluaqutaanngilatit...Men du er ikke til nogen gavn for dine stjerner...

§BEN

(Dan.XIII.108-109)- Inersimasut ilumuuik eqqumeeqaat, ataqqinartuaraq silarsuarmut angalanini ingerlateqqileramiuk imminut oqarfigaaq.

De voksne er nu bestemt meget mærkelige, sagde han til sig selv, da han drog videre ud i verden.

§DES-INTO

§REFL

XIV (Dan.XIV.3)Aqqusernup qullianut qullermillu ikitsissummut taamaallaat inissaqaqqatarpoq.

Der var lige akkurat plads til, at der kunne stå en lygtepæl og en lygtetænder.

§TH + conjunction

(Dan.XIV.4)Ataqqinartuaqqap paasisinnaanngilaa aqqusernup qullia qullermillu ikitsissut sumut iluaqutaanersut tamaani qilammi ullorissami illoqaranilu inoqanngitsuinnarsuarmi.

Den lille prins kunne slet ikke forstå, hvad en lygtepæl og en lygtetænder havde at gøre helt her oppe i himmelen på en planet uden huse og uden beboere.

§CAU

(Dan.XIV.13)Ataqqinartuaraq ullorissamut apuukkami qullermik ikitsisartoq ataqqinnilluni ilassivaa.

Da han kom til planeten, hilste han ærbødigt på lygtetænderen.

§DES-GOAL

(Dan.XIV.18)- Ilinnut malittarisassat qanoq ittuugamik? Hvad går dine instrukser ud på? §REC

(Dan.XIV.40)

38

Liv Molich



Minutsimut ataasiarlunga qulleq ikillugulu qamittarpara.

Så har jeg ikke et øjebliks ro, nu da den drejer en omgang i minuttet.

§FREQ

(Dan.XIV.52)Ataqqinartuaqqap qullermik ikitsisartoq isigimmersorpaa, taannalu malittarisassanut naalattorujussuaq nuannarilertorujussuuaa.

Den lille prins stod og så på ham, og han kom til at holde så meget af dennelygtetænder, der var så tro mod sine instrukser.

§TH

(Dan.XIV.68)

Angut taanna, silarsuarmut ingerlaqqiutigaluni ataqqinartuaraq imminutoqarfigaaq, taanna angut angutit sinnerisa tamarmik ammut isigissavaat.

Denne mand, sagde den lille prins til sig selv, medens han drog videre ud i verden, vil alle de andre se ned på, både kongen og den forfængelige, drankeren og forretningsmanden.

§DES-INTO

§REFL§DES-SEM

(Dan.XIV.73)Ulloriaali mikivallaarpoq, inunnummi marlunnut inissaqanngilaq ...

Men hans planet er faktisk for lille.

(Dan.XIV.74)Der er ikke plads til to... §CONT(Dan.XIV.75)

Ataqqinartuaqqalli imminut nassuerutigiumanngisaa tassaavoq ulloriaaraq taanna ungaginerpaajugamiuk, pingaartumik pissutigalugu tassani seqineq ullormut ettuusintfoororiarluni tarrittarmat.

Men hvad den lille prins ikke turde indrømme over for sig selv, det var at han længtes efter lygtetænderens velsignede lille planet, navnlig på grund af dens fjorten hundrede og fyrretyve solnedgange i døgnet.

§REFL

§FREQ

XV (Dan.XV.6)Ataqqinartuaraq nerrivimmut qaqivoq, ingillunilu anersaartulaaqqaarallarpoq.

Den lille prins satte sig op på bordet og pustede lidt.

§DES-ONTO

(Dan.XV.33)Tassami nunalerutilerisoq tamani tamaani angalaarujoornissaminut pingaaruteqarpallaaqaaq.

Geografen er alt for betydelig en mandtil at strejfe om.

§COMP

(Dan.XV.45)- Taava nalunngisaqarpunga ilisimasassarsiortuunissaminut tulluanngilluinnartumik, ataqqinartuaraq oqarpoq.

Så kender jeg en, der ville være en meget dårlig opdagelsesrejsende, sagde den lille prins.

§TP

(Dan.XV.78)Uagut naassaanngitsumut piusartut allattortarpagut.

Vi skriver om de evige ting… §TP

39

Liv Molich



(Dan.XV.81)- Innermik anitsisartut anitsineq ajorpata anitsisaraluarpataluunniit uagutsinnut soqutaanngilaq, nunalerutilerisoq oqarpoq.

Om det er ildsprudende eller ej, det er os det samme, sagde geografen.

§COG

(Dan.XV.82)Uagutsinnut pingaarutilik tassaavoq qaqqaq.

For os er det bare bjerget, der tæller. §COG

(Dan.XV.84)- Kisiannimi ataavartussuseqannginneq qanoq isumaqarami? ataqqinartuaraq apereqqippoq, aperereersimagaangamimi akineqarnissami tungaanut tunniutiinnanngisaannarami.

Men hvad betyder efemerisk? §DES-SEM

(Dan.XV.85)spurgte den lille prins, som aldrig i sit liv havde ladet et spørgsmål ligge, når han een gang havde stillet det.(Dan.XV.89)

Naasuutiga ataavartussuseqanngilaq, ataqqinartuaraq imminut oqarfigaaq, sisamaannarnillu kapinartoqarpoq silarsuarmut imminut illersuutigisinnaasaminik!

Min blomst er efemerisk, sagde den lille prins til sig selv, og den har kun fire torne at forsvare sig mod verden med!

§REFL

§CAU§REFL

40

in west greenlandic...department of language, literature and media institute of culture, language...

Documents